Anomaly Detection in Internet of Things Using Weighted Botox Optimization Algorithm with Gated Recurrent Unit

Anomaly Detection (AD) play a significant part in risk management and data security among various domains like medical image recognition, economic insurance security, Internet of Things (IoT) management. However, several types of AD have been established in traditional research that concentrate on identifying anomaly with the help of classification issues. To overcome this problem, Weighted Botox Optimization Algorithm with Gated Recurrent Unit (WBOA-GRU) is proposed in this research for AD. The WBOA chooses optimal features and helps to reduce dimensionality through updating their locations based on its behavior. The GRU effectively handle sequential data when requiring less computational sources thereby enhancing the classification accuracy. The preprocessing uses two methods where hash encoding is applied to convert categorical data into numerical format and then normalization is applied to scale the data for ensuing reliability. The WBOA-GRU achieves accuracy of 99.25% for UNSW-NB15 dataset which is better than Rat Swarm Optimization with Improved Self-Attention based Gated Recurrent Unit (RSO-ISAGRU).